In this study, 140 coronary patients have undergone comprehensive evaluations which include exercise stress testing, 24 hour dietary recall, fasting lipid profiles, insulin and glucose levels and dual x-ray absorptiometry. Of these patients, 97 have returned for repeat testing after 3 months of exercise rehabilitation. The primary study hypotheses are:1. Baseline measures of body composition and/or body fat distribution, will predict baseline values of coronary risk factors such as serum triglycerides, HDL-cholesterol, Peak aerobic capacity and fasting insulin levels2. Training induced changes in body composition and/or body fat distribution will predict changes in measures of coronary risk factors such as triglycerides, HDL-cholesterol and fasting insulin levels. Our initial analysis showed that baseline body fat distribution, as measured by the waist to hip ratio (WHR), was the best predictor of serum triglyceride levels, HDL-cholesterol and fasting insulin levels in older women with coronary artery disease, compared with a number of other clinically available measures such as dietary intake of macronutrients. On the other hand, peak aerobic capacity was the best predictor of LDL- cholesterol. We will soon be reanalyzing our data on baseline predictors to include both genders and the entire age spectrum. In addition, we will analyze effects of exercise training on body fat distribution and related coronary risk factors.